Top Stories from the Microsoft DevOps Community – 2019.08.30
Sasha has been out at DevOpsDays Chicago this week so my turn to pick the highlights from the amazing Azure DevOps community across my feeds. Worth mentioning that while DevOpsDays Chicago has finished, there are plenty of local DevOpsDays happening across the globe. Next week it’s the turn of Cape Town where Rory Preddy will be showing you how to integrated Accessibility Insights into your CI/CD pipeline in Azure. Remember, DevOps is about improving the flow of value to all your end users not just the temporarily unimpaired. Building accessibility into your pipelines is a great way to make sure your team see it as adding value rather than make it feel like just a compliance checkbox for them.
The irrepressible Anthony Shaw has updated his fantastic PyTest extension that adds rich test visualization capabilities to Azure Pipelines for Python users. The latest update brings lots of improvements into the test report. You now see the rest run, the time taken for each test module and for each specific test. Docstring support has been improved – as well allowing you to see what the test module does and you get to see docstrings for the test function inside the UI.Parametrized test cases cloned from the properties of the host test function. If you supply a reason when using the pytest.mark.skipit marker that that is also passed through into the report. Any keyword filters, marker filters or file path filters are shown along with the actual command used to run pytest.
If you are doing Python projects, when testing them you will often be using Virtual Environment and you might have struggled getting them to work in your CI/CD Pipelines. In this short post, Dirk Avery explains how to do this correctly inside of Azure Pipelines.
While we work well with Jira and you can have excellent traceability from Azure Pipelines through something like Jenkins and into Jira, we’re seeing more and more teams looking at migrating from Jira into Azure Boards. Chevron have an excellent case study on their experiences doing this here. Peter Rombouts has published an interesting post documenting his teams experiences moving from Jira to Azure DevOps while making use of the excelent (and open source) Jira to Azure DevOps work item migration tool from the clever folks at Solidify.
We’ve all seen the basic demos on how to create docker images from Azure Pipelines – it’s easy right? Just run docker build & docker push and you are done? What about when you are building the Docker images used for 7 versions of Linux across 3 different distributions, 4 versions of Windows, 3 supported versions of your software and support for x64, AMD32 + AMD64? Soon you end up with over 100 images with over 300 tags. That’s exactly the challenge faced by the .NET Team. Matt Thalman has a fantastic post where he explains how his team create the Docker images for .NET – and the great part is that their YAML pipeline is all open source and available on GitHub so you can follow along in code with Matt’s excellent post as a guide. (Spoiler alert: the matrix tag saves the day again)
Christof Van Geendertaelen shares a nice approach he took to create a easily maintainable administration UI where we has separated responsibility of concerns of the UI from the data it uses and also created an Azure Pipeline to ensure maintenance of the script is done using CI/CD. I’m personally always feeling like I’m dabbling with PowerShell so it’s great to learn more from the experts, and this is a really interesting approach that hopefully really sets up his customer for success in maintaining the script after he has left.
Anyone that knows me knows that I have a penchant for IoT. From Brian the Build Bunny to my Raspberry Pi powered pumpkin to Das Deployer. I love hooking up the real world into the DevOps pipeline. However doing these types of projects barely qualifies as IoT as the Raspberry Pi’s I tend to use are such powerful ARM based devices. Dan Benitah has been posting on Hackster.io about building a pipeline to deploy to his Particle based devices and it’s worth a read for a real example of IoT deployments.
Finally – we’re about to hit the fall conference season, as well as DevOpsDays events you’ll see folks from the Azure DevOps team at a load of events including WinOps, All Things Open, Code.Talks, All Day DevOps, Microsoft Ignite and GitHub Universe. Feel free to stop them and say ‘Hi’ – we’re often known to have stickers with us!
If you’ve written an article about Azure DevOps or find some great content about DevOps on Azure, please share it with us using the #AzureDevOps hashtag on Twitter!